The Agent Skills Directory

[PROMPT_INJECTION]: The instructions contain strong behavioral directives to maintain the stylistic mode across multiple turns. These are character-focused constraints and do not attempt to override the underlying agent safety guidelines or extract internal configurations.
[SAFE]: The skill proactively includes an 'Auto-Clarity' section that mandates switching back to normal language for critical security contexts, such as security warnings and confirmations for irreversible actions.
[SAFE]: No external network requests, sensitive file access, or obfuscated payloads were detected in the skill instructions.

caveman